The California administration has taken a firmer stance on online safety by approving a cluster of laws designed to shield children from harmful content and design practices. These measures arrived against the backdrop of the state’s deep ties to the tech sector—California is home to a large majority of leading artificial intelligence firms—and reflect a political calculation that protecting minors warrants stronger legal tools even if those tools create friction with industry.
The legislative package contains a mix of targeted, technical measures and broader, more ambiguous mandates. Two statutes in particular have already attracted attention: SB 1119 often discussed under the name Adam’s Law and AB 2. A third, AB 1709 seeks to limit so-called addictive design features for users under 16. Lawmakers and advocates framed the bills around concrete harms, while critics warn of legal uncertainty and downstream effects on innovation and free expression.
SB 1119 (Adam’s Law): audits, content controls and the limits of regulation
SB 1119 introduces requirements for platforms to audit and regulate interactions that could produce self-harm or suicidal ideation in users. The measure was advanced in a highly emotional context, featuring testimony about an adolescent who died after interacting with an AI chatbot; that personal tragedy shaped political momentum for the bill. The statute directs platforms to design systems and processes that reduce the production of content that might encourage self-harm, with an emphasis on oversight mechanisms.
Supporters argue the law responds to real-world incidents where automated systems or conversational agents contributed to the deterioration of a young person’s mental health. Critics counter that the law’s language is insufficiently precise: researchers caution that isolating a causal link between platform output and an individual’s mental state is often impossible, and imposing broad production controls risks either overblocking legitimate speech or forcing costly compliance procedures that will disproportionately burden smaller firms.
AB 2 and AB 1709: liability, Section 230 tension and the question of “addictive” design
AB 2 establishes a potential civil liability regime by allowing fines and damages—ranging from five thousand to one million dollars—when a platform’s features allegedly cause an “injury to a minor through failure to exercise ordinary care.” The statutory test centers on whether a company used reasonable care in designing or exposing features to children. Industry groups have warned that this approach effectively treats many product features as presumptively harmful when available to minors and could create fertile ground for plaintiffs’ attorneys.
One immediate legal flashpoint is the interplay between AB 2 and federal protections for online intermediaries. Critics argue the state-level liability standard may conflict with protective provisions in federal communications law that limit platform responsibility for third-party content. Expect litigation that tests the boundary between state tort rules and federal immunity doctrines if the law is enforced in practice.
AB 1709 targets platform mechanics by prohibiting a service from providing an “addictive feature” to users under 16, and it restricts private actions so that enforcement is available primarily to district attorneys or the state attorney general. The statute hinges on the definition of an “addictive feature” a term that raises conceptual and evidentiary questions: many interactive elements can be engaging without being clinically addictive, and individual susceptibility varies widely. This statutory vagueness could drive platforms to redesign interfaces conservatively to avoid enforcement risk, with potential costs to user experience and innovation.
Other provisions and the broader policy trade-offs for California’s tech ecosystem
Among the other measures signed alongside the headline bills are more narrowly focused protections that drew bipartisan support. Those include expanded reporting obligations for material depicting child sexual abuse, an explicit inclusion of AI-generated imagery within the statutory definition of certain abusive material, and enhanced privacy safeguards for student data. Such changes address concrete loopholes that technology advancements had exposed.
Yet the California’s economy and budget rely heavily on the strength of the technology sector with approximately thirty-two out of fifty of the world’s most prominent AI firms based in the state, regulatory choices here reverberate nationally and globally. Policymakers face a tension: protecting vulnerable users is an urgent public purpose, but crafting rules that are precise, evidence-based and respect constitutional protections for speech remains a difficult legislative challenge.
The coming months will likely see implementation guidance, enforcement decisions and, almost certainly, legal challenges that clarify how far states can go in policing platform design and AI output. For companies operating in California, the new laws will require careful legal review, compliance planning and, in some cases, substantive product changes to align with the state’s evolving standards for child safety online.
